Hall et al. BMC Medicine 2014, 12:103 http://www.biomedcentral.com/1741-7015/12/103
RESEARCH ARTICLE
Open Access
Effects of palmitate on genome-wide mRNA expression and DNA methylation patterns in human pancreatic islets Elin Hall1, Petr Volkov1, Tasnim Dayeh1, Karl Bacos1, Tina Rönn1, Marloes Dekker Nitert2 and Charlotte Ling1*
Abstract Background: Circulating free fatty acids are often elevated in patients with type 2 diabetes (T2D) and obese individuals. Chronic exposure to high levels of saturated fatty acids has detrimental effects on islet function and insulin secretion. Altered gene expression and epigenetics may contribute to T2D and obesity. However, there is limited information on whether fatty acids alter the genome-wide transcriptome profile in conjunction with DNA methylation patterns in human pancreatic islets. To dissect the molecular mechanisms linking lipotoxicity to impaired insulin secretion, we investigated the effects of a 48 h palmitate treatment in vitro on genome-wide mRNA expression and DNA methylation patterns in human pancreatic islets. Methods: Genome-wide mRNA expression was analyzed using Affymetrix GeneChip® Human Gene 1.0 ST whole transcript-based array (n = 13) and genome-wide DNA methylation was analyzed using Infinium HumanMethylation450K BeadChip (n = 13) in human pancreatic islets exposed to palmitate or control media for 48 h. A non-parametric paired Wilcoxon statistical test was used to analyze mRNA expression. Apoptosis was measured using Apo-ONE® Homogeneous Caspase-3/7 Assay (n = 4). Results: While glucose-stimulated insulin secretion was decreased, there was no significant effect on apoptosis in human islets exposed to palmitate. We identified 1,860 differentially expressed genes in palmitate-treated human islets. These include candidate genes for T2D, such as TCF7L2, GLIS3, HNF1B and SLC30A8. Additionally, genes in glycolysis/ gluconeogenesis, pyruvate metabolism, fatty acid metabolism, glutathione metabolism and one carbon pool by folate were differentially expressed in palmitate-treated human islets. Palmitate treatment altered the global DNA methylation level and DNA methylation levels of CpG island shelves and shores, 5′UTR, 3′UTR and gene body regions in human islets. Moreover, 290 genes with differential expression had a corresponding change in DNA methylation, for example, TCF7L2 and GLIS3. Importantly, out of the genes differentially expressed due to palmitate treatment in human islets, 67 were also associated with BMI and 37 were differentially expressed in islets from T2D patients. Conclusion: Our study demonstrates that palmitate treatment of human pancreatic islets gives rise to epigenetic modifications that together with altered gene expression may contribute to impaired insulin secretion and T2D. Keywords: Palmitate, Human pancreatic islets, Type 2 diabetes, Lipotoxicity, DNA methylation, mRNA expression, Insulin secretion, Epigenetics
* Correspondence: [email protected] 1 Epigenetics and Diabetes Unit, Department of Clinical Sciences, Lund University Diabetes Centre, CRC, Lund University, Scania University Hospital, Malmö, Sweden Full list of author information is available at the end of the article © 2014 Hall et al.; licensee BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/4.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly credited.
Hall et al. BMC Medicine 2014, 12:103 http://www.biomedcentral.com/1741-7015/12/103
Background The risk of developing type 2 diabetes (T2D) is influenced by both genetic and environmental factors. While genome-wide association studies (GWAS) have identified more than 60 single nucleotide polymorphisms (SNPs) associated with an increased risk for T2D [1,2], obesity, physical inactivity and ageing represent non-genetic risk factors for the disease. Recent studies suggest that epigenetic factors, such as DNA methylation, play a role in the pathogenesis of T2D [3-11]. Nevertheless, genome-wide human epigenetic studies linking altered DNA methylation to diabetes remain scarce. In mammalian cells DNA methylation mainly occurs at the cytosine of CpG dinucleotides. Methylated CpG sites can alter transcriptional activity by interfering with binding of transcription factors in promoter regions or by recruiting methyl binding proteins which in turn may recruit histone deacethylases and transcriptional co-repressors [3]. Increased DNA methylation of beta-cell specific genes, such as PDX-1 and INS, correlates negatively with the expression of respective genes in pancreatic islets from T2D patients [4,5]. Plasma levels of free fatty acids are often elevated in T2D patients and in obese individuals [12,13]. Chronic exposure to high levels of fatty acids has negative effects on beta-cell function [12,13]. The severity of this effect depends on the length and saturation of fatty acids. Long chain saturated fatty acids, for example, palmitate and stearate, are reportedly more cytotoxic than the long chain unsaturated fatty acid oleate [14-16], and long term treatment (≥48 h) with palmitate reduces glucose-stimulated insulin secretion in rodent islets and clonal beta-cells [17,18]. Moreover, prolonged exposure to non-esterified fatty acids in vivo also resulted in impaired islets function and decreased glucosestimulated insulin secretion in humans [19,20]. Additionally, transcriptome analyses of clonal beta-cells revealed differences in the gene expression pattern in cells treated with high palmitate concentrations. Specifically, palmitate exposure altered the expression of genes with a role in fatty acid metabolism and steroid biosynthesis [21,22]. In clonal beta-cells, palmitate exposure also altered histone modifications [22]. As most of the cell types in pancreatic islets affect whole body energy homeostasis [23], it is essential to also study the impact of fatty acids on intact human islets. However, while some studies have analyzed expression of specific genes in human islets exposed to palmitate [24-27], to our knowledge no previous study has analyzed the genome-wide expression profile in palmitate-treated human islets of more than five human donors [28,29]. Moreover, whether the genome-wide DNA methylation pattern is affected by fatty acids in human islets remains unknown. The aim of this study was therefore to investigate if treatment with palmitate for 48 h affects genome-wide mRNA expression and DNA methylation patterns in
Page 2 of 15
human pancreatic islets and, consequently influences glucose-stimulated insulin secretion and/or apoptosis. To validate our in vitro findings, we related genomewide gene expression in human islets to BMI in nondiabetic individuals and to T2D in a case-control cohort.
Methods Human pancreatic islets
Pancreatic islets from 13 donors were included in the genome-wide RNA and DNA methylation array analyses. While pancreatic islets from eight donors were included in both the mRNA array analysis and the DNA methylation array analysis, pancreatic islets from five donors were unique for each array (Table 1 and Additional file 1: Table S1). The impact of body mass index (BMI) on gene expression was studied in pancreatic islets from 87 nondiabetic donors (53 males and 34 females, BMI ranged between 17.6 to 40.1 kg/m2, mean BMI = 25.8 ± 3.4 kg/m2, age = 56.7 ± 10.5 years). The impact of T2D on gene expression was studied in pancreatic islets from 15 donors (10 males and 5 females, age = 59.5 ± 10.7 years and mean BMI = 28.3 ± 4.7 kg/m2) diagnosed with T2D and 34 nondiabetic donors (22 males and 12 females, age = 56.0 ± 9.0 years and mean BMI = 28.3 ± 4.7 kg/m2) with an HbA1c below 6.0%. Informed consent for organ donation for medical research was obtained from pancreatic donors or their relatives in accordance with the approval by the regional ethics committee in Lund, Sweden (Dnr 173/ 2007). This study was performed in agreement with the Helsinki Declaration. Human pancreatic islets were prepared by collagenase digestion and density gradient purification. The islet purity was 80% ± 2.5%, as assessed by the ratio of expression of islet (INS, GCG, and SST) and non-islet specific (AMY2A, PNLIP, CTRC) genes. Preparation of medium containing palmitate
First, a stock solution of 10 mM palmitate and 10% fatty acid free BSA was created. A total of 128 mg palmitate was dissolved in 50 ml 99% ethanol and then 60 μl 10 M NaOH was added. The solution was vacuum-dried and then resolved in 25 ml H2O during heating. Next, 6 g of fatty acid free BSA was dissolved in 24 ml H2O and then Table 1 Characteristics of human pancreatic donors included in the mRNA expression array analysis n (male/female)
13 (7/6)
Age (years)
55 ± 14
BMI (kg/m2)
25.5 ± 4.3
HbA1c* (%)
5.6 ± 0.9
HbA1c* (mmol/mol)
47.6 ± 9.2
Data are expressed as mean ± SD. BMI, Body Mass Index. *Data available for eight donors (five males and three females).
Hall et al. BMC Medicine 2014, 12:103 http://www.biomedcentral.com/1741-7015/12/103
25 ml was taken and mixed with the 25 ml palmitate solution. The stock solution was then diluted to a final concentration of 1 mM palmitate and 1 weight % BSA (corresponding to 0.15 mM BSA) in the CMRL 1066 medium (ICN Biomedicals, Costa Mesa, CA, USA) supplemented with 10 mM nicotinamide (Sigma-Aldrich, Sweden, Stockholm), 10 mM HEPES buffer (GIBCO, BRL, Gaithersburg, MD, USA), 0.25 μg/ml fungizone (GIBCO), 50 μg/ml gentamicin, 2 mM L-glutamine (GIBCO), 10 μg/ml Ciprofloxacin (Bayer Healthcare, Leverkusen, Germany), 10% (v/v) heat-inactivated human serum and 5.56 mM glucose. The molar (mmol/l) ratio of palmitate/BSA concentrations was 6.6:1 in the culture medium. Palmitate treatment
To study the impact of palmitate-induced lipotoxicity on human islets, approximately 1,000 islets from each donor (n = 13) were cultured for 48 h in CMRL 1066 medium (including 5.56 mM glucose) either with (lipotox) or without (control) 1 mM palmitate conjugated with 1% BSA (corresponding to 0.15 mM BSA) (Figure 1a). The same treatment time and palmitate/BSA ratio have been used in previous studies examining the impact of lipotoxicity on islet function and was therefore selected in the present study [22,30]. Circulating non-esterified fatty acid levels have been reported to range between 0.59 to 0.83 mM for overweight, non-diabetic individuals (BMI approximately 26 kg/m2) and between 0.69 to 0.975 mM for overweight, diabetic individuals (BMI of approximately 29 kg/m2) [31]. The 1 mM palmitate used in the current study, which is close to the upper limit of the reference range, mimics the levels reported in overweight/obese individuals with diabetes. After 48 h DNA and RNA were extracted, glucose-stimulated insulin secretion was analyzed and/or apoptosis assays were performed. RNA and DNA isolation
DNA and RNA were extracted from the human pancreatic islets using the AllPrep DNA/RNA kit (Qiagen, Hilden, Germany) according to the manufacturer’s instructions. RNA quantity and quality were assessed by Nanodrop (Nanodrop, Wilmington, DE, USA). The 260/280 ratios of all samples were between 1.98 and 2.16. The integrity and quality of the RNA was assessed using the Bioanalyzer (Agilent Technologies, Santa Clara, CA, USA). All RNA integrity number (RIN) values were ≥7.4. Microarray mRNA expression analysis
The Affymetrix GeneChip® Human Gene 1.0 ST whole transcript-based array (Affymetrix, Santa Clara, CA, USA) covering 28,869 genes was used to analyze mRNA expression (Figure 1b) in pancreatic islets from 13 human
Page 3 of 15
donors (Table 1) exposed to palmitate or control conditions (in total 26 samples) and in pancreatic islets from 87 non-diabetic donors as well as from diabetic and nondiabetic donors, according to the manufacturer’s recommendations. The Oligo package from Bioconductor was used to compute Robust Multichip Average expression measures [32]. Genome-wide DNA methylation analysis
A total of 500 ng genomic DNA from human pancreatic islets of 13 donors (Additional file 1: Table S1) exposed to palmitate or control conditions (in total 26 samples) was bisulfite-converted with the EZ DNA methylation kit (Zymo Research Corporation, Irvine, CA, USA). DNA methylation was analyzed by using the Infinium HumanMethylation450K BeadChip (Illumina, San Diego, CA, USA) which contains 485,577 probes and covers 99% of all RefSeq genes [33]. Bisulfite converted DNA was used to analyze DNA methylation with the Infinium® assay according to the standard Infinium HD Assay Methylation Protocol (Part # 15019519, Illumina). The Infinium HumanMethylation450K BeadChips were then imaged with the Illumina iScan. The raw methylation score for each CpG site, which is represented as β-value, was calculated using the GenomeStudio® methylation module software. The β-values were calculated as (β = intensity of the methylated allele (M)/(intensity of the Unmethylated allele (U) + intensity of the Methylated allele (M) + 100)). All samples passed GenomeStudio® quality control steps based on built-in control probes for staining, hybridization, extension and specificity, and displayed high quality bisulfite conversion efficiency with an intensity signal above 4,000 [34]. Probes were filtered away from further analysis based on a mean detection P-value >0.01. After quality control analysis, DNA methylation data were obtained for 483,844 probes. β-values were then converted to M-values (M = log2 (β/(1 - β))) for further bioinformatic and statistical analyses of the methylation data [35]. Background and quantile normalization was performed using the lumi package from Bioconductor [36]. Background correction was performed by subtracting the median M-value of the 600 built-in negative controls and methylation data were further normalized using quantile normalization [37]. ComBat was used to adjust for batch effects between arrays [38]. A linear regression model was used to identify differences in DNA methylation between control and palmitate-treated islets in a paired fashion as described elsewhere [39]. As β-values are biologically easier to interpret, M-values were reconverted to β-values when describing the DNA methylation results. The DNA methylation probes on the Infinium HumanMethylation450K BeadChip have been annotated to different genomic regions depending on their location in relation to a gene or a CpG island [33].
Hall et al. BMC Medicine 2014, 12:103 http://www.biomedcentral.com/1741-7015/12/103
Page 4 of 15
Figure 1 Study design and work flow. Study design for the lipotoxicity study in human pancreatic islets is presented in panel a Work flow for the analysis of mRNA expression data in combination with DNA methylation data in human pancreatic islets exposed to palmitate is presented in panel b.
KEGG pathway analysis
Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway analysis of expression data was performed with the online tool WebGestalt [40,41] (accessed 27 March 2012 and 12 February 2014). For the pathway analysis of mRNA expression data, Affymetrix probe IDs were used to identify unique genes and Affymetrix
GeneChip® Human Gene 1.0 ST genes were used as background in this analysis. For the pathway analysis of the DNA methylation data, the gene symbol was used to identify unique genes and the human genome was used as background in this analysis. The Benjamini and Hochberg method was used to correct P-values for multiple testing.
Hall et al. BMC Medicine 2014, 12:103 http://www.biomedcentral.com/1741-7015/12/103
Glucose-stimulated insulin secretion
Glucose-stimulated insulin secretion was analyzed in control and palmitate-treated human islets from nine donors. After 48 h culture in control or palmitate-containing medium, 10 replicates of 10 human islets per culture condition (control and palmitate-treated) and donor were pre-incubated in HEPES-balanced salt solution (HBSS) containing (in mM) 114 NaCl, 4.7 KCl, 1.2 KH2PO4, 1.16 MgSO4, 20 HEPES, 25.5 NaHCO3, 2.5 CaCl2 at pH 7.2 with 0.575 BSA and 3.3 mM glucose (1.65 mM glucose for one sample) for 1 h at 37°C. Thereafter, for each donor, glucose was added to five of the replicates to a final concentration of 16.7 mM glucose (15.05 mM glucose for one sample) to study glucose-stimulated insulin secretion and the other five replicates were kept in 3.3 mM glucose to study basal insulin secretion and the incubation was continued for one more hour. The supernatant was immediately removed and the insulin concentration in the medium was measured by radioimmunoassay (RIA) (Millipore, Uppsala, Sweden).
Page 5 of 15
as mean ± standard error of mean (sem), unless stated otherwise.
Results Impaired insulin secretion in human islets exposed to palmitate
To investigate the physiological response to 1 mM palmitate treatment for 48 h, we measured glucose-stimulated insulin secretion in human islets cultured under control (5.56 mM glucose) or lipotoxic (5.56 mM glucose and 1 mM palmitate) conditions. We found decreased glucosestimulated insulin secretion measured as fold change (insulin secretion at high glucose levels/insulin secretion at low glucose levels) in the palmitate-treated compared with control-treated human islets (Figure 2a). We also evaluated the effect of the palmitate treatment on apoptosis in human islets by measuring the combined activity of Caspase-3 and -7. Palmitate treatment did not alter islet cell apoptosis rates (P = 0.62, Figure 2b).
Assessment of apoptosis in human pancreatic islets
Apoptosis was measured in islets from four human donors with the Apo-ONE® Homogeneous Caspase-3/7 Assay (Promega, Madison, WI, USA) as described elsewhere [42]. The assay contains proflourescent rhodamin 110 (Z-DEVD-R110) which serves as a substrate for both Caspase-3 and -7. Upon lysis of cells, the available Caspase -3/-7 in the sample will cleave Z-DEVD-R110 to fluorescent rhodamine 110, which is then measured. Subsequently, the assay measures the combined activity of Caspase-3 and -7. After 48 h incubation in control or palmitate medium, triplicates of 20 human pancreatic islets each were handpicked from each culture condition, washed and transferred to a plate containing HBSS. After 1.5 h, fluorescence was measured with a Tecan Infinite M200pro plate reader (Tecan Group Ltd., Männedorf, Switzerland) to determine the Caspase-3/7 activity. Statistics
A non-parametric paired test (Wilcoxon) was used to identify differences in mRNA expression between control and palmitate-treated human islets. A False Discovery Rate (FDR) analysis was performed to correct for multiple testing in the mRNA expression data. Genes exhibiting differential expression with a FDR below 5% (q
RESEARCH ARTICLE
Open Access
Effects of palmitate on genome-wide mRNA expression and DNA methylation patterns in human pancreatic islets Elin Hall1, Petr Volkov1, Tasnim Dayeh1, Karl Bacos1, Tina Rönn1, Marloes Dekker Nitert2 and Charlotte Ling1*
Abstract Background: Circulating free fatty acids are often elevated in patients with type 2 diabetes (T2D) and obese individuals. Chronic exposure to high levels of saturated fatty acids has detrimental effects on islet function and insulin secretion. Altered gene expression and epigenetics may contribute to T2D and obesity. However, there is limited information on whether fatty acids alter the genome-wide transcriptome profile in conjunction with DNA methylation patterns in human pancreatic islets. To dissect the molecular mechanisms linking lipotoxicity to impaired insulin secretion, we investigated the effects of a 48 h palmitate treatment in vitro on genome-wide mRNA expression and DNA methylation patterns in human pancreatic islets. Methods: Genome-wide mRNA expression was analyzed using Affymetrix GeneChip® Human Gene 1.0 ST whole transcript-based array (n = 13) and genome-wide DNA methylation was analyzed using Infinium HumanMethylation450K BeadChip (n = 13) in human pancreatic islets exposed to palmitate or control media for 48 h. A non-parametric paired Wilcoxon statistical test was used to analyze mRNA expression. Apoptosis was measured using Apo-ONE® Homogeneous Caspase-3/7 Assay (n = 4). Results: While glucose-stimulated insulin secretion was decreased, there was no significant effect on apoptosis in human islets exposed to palmitate. We identified 1,860 differentially expressed genes in palmitate-treated human islets. These include candidate genes for T2D, such as TCF7L2, GLIS3, HNF1B and SLC30A8. Additionally, genes in glycolysis/ gluconeogenesis, pyruvate metabolism, fatty acid metabolism, glutathione metabolism and one carbon pool by folate were differentially expressed in palmitate-treated human islets. Palmitate treatment altered the global DNA methylation level and DNA methylation levels of CpG island shelves and shores, 5′UTR, 3′UTR and gene body regions in human islets. Moreover, 290 genes with differential expression had a corresponding change in DNA methylation, for example, TCF7L2 and GLIS3. Importantly, out of the genes differentially expressed due to palmitate treatment in human islets, 67 were also associated with BMI and 37 were differentially expressed in islets from T2D patients. Conclusion: Our study demonstrates that palmitate treatment of human pancreatic islets gives rise to epigenetic modifications that together with altered gene expression may contribute to impaired insulin secretion and T2D. Keywords: Palmitate, Human pancreatic islets, Type 2 diabetes, Lipotoxicity, DNA methylation, mRNA expression, Insulin secretion, Epigenetics
* Correspondence: [email protected] 1 Epigenetics and Diabetes Unit, Department of Clinical Sciences, Lund University Diabetes Centre, CRC, Lund University, Scania University Hospital, Malmö, Sweden Full list of author information is available at the end of the article © 2014 Hall et al.; licensee BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/4.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly credited.
Hall et al. BMC Medicine 2014, 12:103 http://www.biomedcentral.com/1741-7015/12/103
Background The risk of developing type 2 diabetes (T2D) is influenced by both genetic and environmental factors. While genome-wide association studies (GWAS) have identified more than 60 single nucleotide polymorphisms (SNPs) associated with an increased risk for T2D [1,2], obesity, physical inactivity and ageing represent non-genetic risk factors for the disease. Recent studies suggest that epigenetic factors, such as DNA methylation, play a role in the pathogenesis of T2D [3-11]. Nevertheless, genome-wide human epigenetic studies linking altered DNA methylation to diabetes remain scarce. In mammalian cells DNA methylation mainly occurs at the cytosine of CpG dinucleotides. Methylated CpG sites can alter transcriptional activity by interfering with binding of transcription factors in promoter regions or by recruiting methyl binding proteins which in turn may recruit histone deacethylases and transcriptional co-repressors [3]. Increased DNA methylation of beta-cell specific genes, such as PDX-1 and INS, correlates negatively with the expression of respective genes in pancreatic islets from T2D patients [4,5]. Plasma levels of free fatty acids are often elevated in T2D patients and in obese individuals [12,13]. Chronic exposure to high levels of fatty acids has negative effects on beta-cell function [12,13]. The severity of this effect depends on the length and saturation of fatty acids. Long chain saturated fatty acids, for example, palmitate and stearate, are reportedly more cytotoxic than the long chain unsaturated fatty acid oleate [14-16], and long term treatment (≥48 h) with palmitate reduces glucose-stimulated insulin secretion in rodent islets and clonal beta-cells [17,18]. Moreover, prolonged exposure to non-esterified fatty acids in vivo also resulted in impaired islets function and decreased glucosestimulated insulin secretion in humans [19,20]. Additionally, transcriptome analyses of clonal beta-cells revealed differences in the gene expression pattern in cells treated with high palmitate concentrations. Specifically, palmitate exposure altered the expression of genes with a role in fatty acid metabolism and steroid biosynthesis [21,22]. In clonal beta-cells, palmitate exposure also altered histone modifications [22]. As most of the cell types in pancreatic islets affect whole body energy homeostasis [23], it is essential to also study the impact of fatty acids on intact human islets. However, while some studies have analyzed expression of specific genes in human islets exposed to palmitate [24-27], to our knowledge no previous study has analyzed the genome-wide expression profile in palmitate-treated human islets of more than five human donors [28,29]. Moreover, whether the genome-wide DNA methylation pattern is affected by fatty acids in human islets remains unknown. The aim of this study was therefore to investigate if treatment with palmitate for 48 h affects genome-wide mRNA expression and DNA methylation patterns in
Page 2 of 15
human pancreatic islets and, consequently influences glucose-stimulated insulin secretion and/or apoptosis. To validate our in vitro findings, we related genomewide gene expression in human islets to BMI in nondiabetic individuals and to T2D in a case-control cohort.
Methods Human pancreatic islets
Pancreatic islets from 13 donors were included in the genome-wide RNA and DNA methylation array analyses. While pancreatic islets from eight donors were included in both the mRNA array analysis and the DNA methylation array analysis, pancreatic islets from five donors were unique for each array (Table 1 and Additional file 1: Table S1). The impact of body mass index (BMI) on gene expression was studied in pancreatic islets from 87 nondiabetic donors (53 males and 34 females, BMI ranged between 17.6 to 40.1 kg/m2, mean BMI = 25.8 ± 3.4 kg/m2, age = 56.7 ± 10.5 years). The impact of T2D on gene expression was studied in pancreatic islets from 15 donors (10 males and 5 females, age = 59.5 ± 10.7 years and mean BMI = 28.3 ± 4.7 kg/m2) diagnosed with T2D and 34 nondiabetic donors (22 males and 12 females, age = 56.0 ± 9.0 years and mean BMI = 28.3 ± 4.7 kg/m2) with an HbA1c below 6.0%. Informed consent for organ donation for medical research was obtained from pancreatic donors or their relatives in accordance with the approval by the regional ethics committee in Lund, Sweden (Dnr 173/ 2007). This study was performed in agreement with the Helsinki Declaration. Human pancreatic islets were prepared by collagenase digestion and density gradient purification. The islet purity was 80% ± 2.5%, as assessed by the ratio of expression of islet (INS, GCG, and SST) and non-islet specific (AMY2A, PNLIP, CTRC) genes. Preparation of medium containing palmitate
First, a stock solution of 10 mM palmitate and 10% fatty acid free BSA was created. A total of 128 mg palmitate was dissolved in 50 ml 99% ethanol and then 60 μl 10 M NaOH was added. The solution was vacuum-dried and then resolved in 25 ml H2O during heating. Next, 6 g of fatty acid free BSA was dissolved in 24 ml H2O and then Table 1 Characteristics of human pancreatic donors included in the mRNA expression array analysis n (male/female)
13 (7/6)
Age (years)
55 ± 14
BMI (kg/m2)
25.5 ± 4.3
HbA1c* (%)
5.6 ± 0.9
HbA1c* (mmol/mol)
47.6 ± 9.2
Data are expressed as mean ± SD. BMI, Body Mass Index. *Data available for eight donors (five males and three females).
Hall et al. BMC Medicine 2014, 12:103 http://www.biomedcentral.com/1741-7015/12/103
25 ml was taken and mixed with the 25 ml palmitate solution. The stock solution was then diluted to a final concentration of 1 mM palmitate and 1 weight % BSA (corresponding to 0.15 mM BSA) in the CMRL 1066 medium (ICN Biomedicals, Costa Mesa, CA, USA) supplemented with 10 mM nicotinamide (Sigma-Aldrich, Sweden, Stockholm), 10 mM HEPES buffer (GIBCO, BRL, Gaithersburg, MD, USA), 0.25 μg/ml fungizone (GIBCO), 50 μg/ml gentamicin, 2 mM L-glutamine (GIBCO), 10 μg/ml Ciprofloxacin (Bayer Healthcare, Leverkusen, Germany), 10% (v/v) heat-inactivated human serum and 5.56 mM glucose. The molar (mmol/l) ratio of palmitate/BSA concentrations was 6.6:1 in the culture medium. Palmitate treatment
To study the impact of palmitate-induced lipotoxicity on human islets, approximately 1,000 islets from each donor (n = 13) were cultured for 48 h in CMRL 1066 medium (including 5.56 mM glucose) either with (lipotox) or without (control) 1 mM palmitate conjugated with 1% BSA (corresponding to 0.15 mM BSA) (Figure 1a). The same treatment time and palmitate/BSA ratio have been used in previous studies examining the impact of lipotoxicity on islet function and was therefore selected in the present study [22,30]. Circulating non-esterified fatty acid levels have been reported to range between 0.59 to 0.83 mM for overweight, non-diabetic individuals (BMI approximately 26 kg/m2) and between 0.69 to 0.975 mM for overweight, diabetic individuals (BMI of approximately 29 kg/m2) [31]. The 1 mM palmitate used in the current study, which is close to the upper limit of the reference range, mimics the levels reported in overweight/obese individuals with diabetes. After 48 h DNA and RNA were extracted, glucose-stimulated insulin secretion was analyzed and/or apoptosis assays were performed. RNA and DNA isolation
DNA and RNA were extracted from the human pancreatic islets using the AllPrep DNA/RNA kit (Qiagen, Hilden, Germany) according to the manufacturer’s instructions. RNA quantity and quality were assessed by Nanodrop (Nanodrop, Wilmington, DE, USA). The 260/280 ratios of all samples were between 1.98 and 2.16. The integrity and quality of the RNA was assessed using the Bioanalyzer (Agilent Technologies, Santa Clara, CA, USA). All RNA integrity number (RIN) values were ≥7.4. Microarray mRNA expression analysis
The Affymetrix GeneChip® Human Gene 1.0 ST whole transcript-based array (Affymetrix, Santa Clara, CA, USA) covering 28,869 genes was used to analyze mRNA expression (Figure 1b) in pancreatic islets from 13 human
Page 3 of 15
donors (Table 1) exposed to palmitate or control conditions (in total 26 samples) and in pancreatic islets from 87 non-diabetic donors as well as from diabetic and nondiabetic donors, according to the manufacturer’s recommendations. The Oligo package from Bioconductor was used to compute Robust Multichip Average expression measures [32]. Genome-wide DNA methylation analysis
A total of 500 ng genomic DNA from human pancreatic islets of 13 donors (Additional file 1: Table S1) exposed to palmitate or control conditions (in total 26 samples) was bisulfite-converted with the EZ DNA methylation kit (Zymo Research Corporation, Irvine, CA, USA). DNA methylation was analyzed by using the Infinium HumanMethylation450K BeadChip (Illumina, San Diego, CA, USA) which contains 485,577 probes and covers 99% of all RefSeq genes [33]. Bisulfite converted DNA was used to analyze DNA methylation with the Infinium® assay according to the standard Infinium HD Assay Methylation Protocol (Part # 15019519, Illumina). The Infinium HumanMethylation450K BeadChips were then imaged with the Illumina iScan. The raw methylation score for each CpG site, which is represented as β-value, was calculated using the GenomeStudio® methylation module software. The β-values were calculated as (β = intensity of the methylated allele (M)/(intensity of the Unmethylated allele (U) + intensity of the Methylated allele (M) + 100)). All samples passed GenomeStudio® quality control steps based on built-in control probes for staining, hybridization, extension and specificity, and displayed high quality bisulfite conversion efficiency with an intensity signal above 4,000 [34]. Probes were filtered away from further analysis based on a mean detection P-value >0.01. After quality control analysis, DNA methylation data were obtained for 483,844 probes. β-values were then converted to M-values (M = log2 (β/(1 - β))) for further bioinformatic and statistical analyses of the methylation data [35]. Background and quantile normalization was performed using the lumi package from Bioconductor [36]. Background correction was performed by subtracting the median M-value of the 600 built-in negative controls and methylation data were further normalized using quantile normalization [37]. ComBat was used to adjust for batch effects between arrays [38]. A linear regression model was used to identify differences in DNA methylation between control and palmitate-treated islets in a paired fashion as described elsewhere [39]. As β-values are biologically easier to interpret, M-values were reconverted to β-values when describing the DNA methylation results. The DNA methylation probes on the Infinium HumanMethylation450K BeadChip have been annotated to different genomic regions depending on their location in relation to a gene or a CpG island [33].
Hall et al. BMC Medicine 2014, 12:103 http://www.biomedcentral.com/1741-7015/12/103
Page 4 of 15
Figure 1 Study design and work flow. Study design for the lipotoxicity study in human pancreatic islets is presented in panel a Work flow for the analysis of mRNA expression data in combination with DNA methylation data in human pancreatic islets exposed to palmitate is presented in panel b.
KEGG pathway analysis
Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway analysis of expression data was performed with the online tool WebGestalt [40,41] (accessed 27 March 2012 and 12 February 2014). For the pathway analysis of mRNA expression data, Affymetrix probe IDs were used to identify unique genes and Affymetrix
GeneChip® Human Gene 1.0 ST genes were used as background in this analysis. For the pathway analysis of the DNA methylation data, the gene symbol was used to identify unique genes and the human genome was used as background in this analysis. The Benjamini and Hochberg method was used to correct P-values for multiple testing.
Hall et al. BMC Medicine 2014, 12:103 http://www.biomedcentral.com/1741-7015/12/103
Glucose-stimulated insulin secretion
Glucose-stimulated insulin secretion was analyzed in control and palmitate-treated human islets from nine donors. After 48 h culture in control or palmitate-containing medium, 10 replicates of 10 human islets per culture condition (control and palmitate-treated) and donor were pre-incubated in HEPES-balanced salt solution (HBSS) containing (in mM) 114 NaCl, 4.7 KCl, 1.2 KH2PO4, 1.16 MgSO4, 20 HEPES, 25.5 NaHCO3, 2.5 CaCl2 at pH 7.2 with 0.575 BSA and 3.3 mM glucose (1.65 mM glucose for one sample) for 1 h at 37°C. Thereafter, for each donor, glucose was added to five of the replicates to a final concentration of 16.7 mM glucose (15.05 mM glucose for one sample) to study glucose-stimulated insulin secretion and the other five replicates were kept in 3.3 mM glucose to study basal insulin secretion and the incubation was continued for one more hour. The supernatant was immediately removed and the insulin concentration in the medium was measured by radioimmunoassay (RIA) (Millipore, Uppsala, Sweden).
Page 5 of 15
as mean ± standard error of mean (sem), unless stated otherwise.
Results Impaired insulin secretion in human islets exposed to palmitate
To investigate the physiological response to 1 mM palmitate treatment for 48 h, we measured glucose-stimulated insulin secretion in human islets cultured under control (5.56 mM glucose) or lipotoxic (5.56 mM glucose and 1 mM palmitate) conditions. We found decreased glucosestimulated insulin secretion measured as fold change (insulin secretion at high glucose levels/insulin secretion at low glucose levels) in the palmitate-treated compared with control-treated human islets (Figure 2a). We also evaluated the effect of the palmitate treatment on apoptosis in human islets by measuring the combined activity of Caspase-3 and -7. Palmitate treatment did not alter islet cell apoptosis rates (P = 0.62, Figure 2b).
Assessment of apoptosis in human pancreatic islets
Apoptosis was measured in islets from four human donors with the Apo-ONE® Homogeneous Caspase-3/7 Assay (Promega, Madison, WI, USA) as described elsewhere [42]. The assay contains proflourescent rhodamin 110 (Z-DEVD-R110) which serves as a substrate for both Caspase-3 and -7. Upon lysis of cells, the available Caspase -3/-7 in the sample will cleave Z-DEVD-R110 to fluorescent rhodamine 110, which is then measured. Subsequently, the assay measures the combined activity of Caspase-3 and -7. After 48 h incubation in control or palmitate medium, triplicates of 20 human pancreatic islets each were handpicked from each culture condition, washed and transferred to a plate containing HBSS. After 1.5 h, fluorescence was measured with a Tecan Infinite M200pro plate reader (Tecan Group Ltd., Männedorf, Switzerland) to determine the Caspase-3/7 activity. Statistics
A non-parametric paired test (Wilcoxon) was used to identify differences in mRNA expression between control and palmitate-treated human islets. A False Discovery Rate (FDR) analysis was performed to correct for multiple testing in the mRNA expression data. Genes exhibiting differential expression with a FDR below 5% (q
